Real-time specified object tracking system under complex background

A new approach is proposed in real-time specified object tracking system under auto dome rotation freely. When it is necessary to track objects in occlusion conditions or with high speed by real-time system, it is not always possible to using particle filter or mean shift algorithm both of which are widely used. The key idea of the presented research work is to integrate the advantages of mean-shift algorithm and particle filter. Using mean shift algorithm to iterate the particles generated by particle filter repeatedly until stable, which could efficiently reduces the number of sampled particles. Therefore the system designed in this paper can track the objects in occlusion conditions or with high speed, and the computational cost is greatly reduced, moreover, an adaptive occlusion detecting algorithm, which update the model when the object is in occlusion makes the system more robust. The experiment results demonstrate that the object tracking system in this paper can meet the tracking requirements when the object is moving, rotating or in partial occlusion. The error of this system is less than 0.859%, it also proved the combing method is effective and steady.

[1]  Dieter Fox,et al.  Adaptive real-time particle filters for robot localization , 2003, 2003 IEEE International Conference on Robotics and Automation (Cat. No.03CH37422).

[2]  Shai Avidan,et al.  Support Vector Tracking , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[3]  Ramin Zabih,et al.  Bayesian multi-camera surveillance , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).

[4]  J.A. Besada,et al.  Design of an A-SMGCS prototype at Barajas airport: airport surveillance sensors bias estimation , 2005, 2005 7th International Conference on Information Fusion.

[5]  Huang Xianwu,et al.  Hand tracking and gesture gecogniton by anisotropic kernel mean shift , 2008, 2008 International Conference on Neural Networks and Signal Processing.

[6]  Jean-Marc Odobez,et al.  Using particles to track varying numbers of interacting people , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[7]  Ramakant Nevatia,et al.  Tracking multiple humans in crowded environment , 2004, Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004..

[8]  Katja Nummiaro.  A Color-based Particle Filter , 2002 .

[9]  Jerry L. Prince,et al.  Fast tracking of cardiac motion using 3D-HARP , 2005, IEEE Transactions on Biomedical Engineering.

[10]  T. Kailath.  The Divergence and Bhattacharyya Distance Measures in Signal Selection , 1967 .

[11]  Tieniu Tan,et al.  Real time hand tracking by combining particle filtering and mean shift , 2004, Sixth IEEE International Conference on Automatic Face and Gesture Recognition, 2004. Proceedings..

[12]  Dorin Comaniciu,et al.  Design, analysis, and engineering of video monitoring systems: an approach and a case study , 2001, Proc. IEEE.

[13]  Takeo Kanade,et al.  Algorithms for cooperative multisensor surveillance , 2001, Proc. IEEE.

[14]  Dorin Comaniciu,et al.  Kernel-Based Object Tracking , 2003, IEEE Trans. Pattern Anal. Mach. Intell..

[15]  Dorin Comaniciu,et al.  Mean Shift: A Robust Approach Toward Feature Space Analysis , 2002, IEEE Trans. Pattern Anal. Mach. Intell..